Hannah Bakker (IOR); Florian Diehlmann (IIP)

Beschreibung:

The objective of Humanitarian Logistics is to provide goods to people in need after the impact of a disaster as fast as possible. However, there is a time lag until resources are available on a large scale since resources are scarce and can only be activated successively. Consequently, decision-makers need to carefully allocate resources to increase the disaster intervention impact. This includes, inter alia, the selection of distribution locations (PoDs) or the allocation of staff to these locations.

In particular, during the early stages of a response to a sudden or gradually evolving crisis, time is a crucial aspect. The number of people in need of support rises steadily while the establishment of a humanitarian supply system requires time. Even though it is crucial to supply as many goods as possible, aspects considering fairness cannot be neglected. One promising approach to include fairness into logistics models is the concept of deprivation levels, which represent an individual’s level of suffering. However, including deprivation costs in logistical models is challenging from modeling, as well as a solution perspective.

Points of distribution (PoDs) for humanitarian goods are capacity constrained, which means that they can only serve a limited number of beneficiaries during a given period, due to, e.g., limited space, staff, and storage options. Most decision support models assume that it is possible to directly assign beneficiaries to certain points of distribution, i.e. through the use of mobile apps, making it possible to control the number of beneficiaries at individual PoDs. However, in reality, such a precise allocation of beneficiaries is oftentimes not feasible and instead, beneficiaries will always choose the closest available PoD. This implies that the number of beneficiaries allocated to a certain PoD is no longer an endogenous modeling decision, but derived as a direct consequence from the locations of the established PoDs. The objective of the Master's thesis is to advance an already existing mathematical program for the PoD installation planning in urban areas to accommodate the idea that beneficiaries will automatically choose the closest location.

The topic requires advancing an already existing mathematical program to account for an additional real-life aspect of a humanitarian relief network. There is case study data available for a catastrophe in Berlin, Germany, and the model should be tested and evaluated based on this case study. Depending on the model extension heuristic methods may be needed to solve the model.

nforderungen: Modellierung, Implementierung, Auswertung

Ansprechpartnerin: Hannah Bakker

Bearbeitungsstatus: (Achtung: nur für Masterarbeiten) frei